rlpp 0.2.5

A GUI tool for building Pygame applications and Reinforcement Learning

Reinforcement Learning with Pygame and Python

This project was inspired by the development of technologies in AI and reinforcement learning applications.
Python has seen the development of very powerful pipelines and libraries like Pytorch and Tensorflow that allow users to create complex projects with deep learning, and therefore gives Python an advantage over perhaps faster but harder to integrate compiled programming languages like C, C# or C++.
However, while Python has the AI libraries, it doesn't have a formal GUI-based library for game-like development like Unity or UnrealEngine does, and while emulating the complexity of building an engine doesn't overcome the benefits of library integration at this point, Python does offer tools for the creation of windows and the coding of agents with the library Pygame.
We hope to narrow the gap between user friendly environment design by providing a simple GUI that works on pyQt5, and offer additional tools to expedite and, to some extent, standardize the structure of Reinforcement Learning Agents that can be deployed in Pygame environments.

This library supplements the online book "The RL Playground with Python", which we consider a good resource for information on the deployment of reinforcement learning agents at an introductory level.
https://ugarcil.github.io/The_RL_Playground_with_python/